Postal voters – and fibbers

Although we talk about polling day, the widespread use of postal voting means that there are now multiple polling days in British general elections. In 2015, postal votes began to be issued around 13 April, as the parties’ manifestos were being published.

As part of the work for The British General Election of 2015, I thought it would be good to know when people actually voted. The fifth wave of the British Election Study has a question specifically asking people if they have a postal vote, and if they have, whether they have used it.

The fifth wave of the BES is a campaign wave, which polls c.600-800 people each day, which allows us to see when people start to say they have used their postal vote. If you just take the % of respondents who say they have voted by post each day during the campaign, you get this graph.

Postal votes

Each data point is a separate poll, which explains why the values can go down as well as up, day-on-day. (Obviously, the cumulative percentage of actual people to have voted by post can only go up, day-on-day).*

The graph shows a clear take off in postal voting from about day 22 of the campaign (that is, 20 April). So postal votes may have started to go out from around 13 April, but by the time they had arrived, and people had prevaricated a bit (left them on the kitchen table, had a ponder, tried to understand the instructions and so on) – it was about nine days later that people started to fill them in and send them back. (And, of course, not all councils will have been quite so organised to get the votes sent out on 13 April). In practice, therefore, most postal voting took place over the last 16 days, or just over two weeks, of the campaign.

The second interesting thing about the graph is all the people who claimed that they had voted by post when it was not simply possible for them to have done so. The percentage saying they had voted by post should be zero for the first 13 days of the campaign, rather than being between 3-6%. It’s a good reminder that some people can fail to recall accurately even the most straight-forward things about their voting.

Philip Cowley

* Technically, this isn’t true: if people were registering to vote at a faster rate than people with postal votes were using them, then the % of people who’d voted by post could decline. The deadline for the former was 20 April, however, around the time people started postal voting in earnest, so we can safely rule it out.